In this example a two-class linear programming machine (LPM) classifier is
trained on a randomly generated examples. The linear programming problem to
which the learning problem is transformed is solved by the CPLEX solver which
must be installed. The LPM regularization parameter is set to C=100 and the bias
term in the classification rule is switched on. The example also shows how to
compute classifier outputs on the training examples and the primal SVM objective
function.

For more details on the LPM see
    Weida Zhou, Li Zhang, Licheng Jiao. Linear programming support vector
    machines. Pattern Recognition, Volume 35, Issue 12, December 2002, 
    pages 2927-2936. 

